Molten-Salt Methane Pyrolysis Optimization Through in-situ Carbon Characterization and

Reactor DesignFabrication & demonstration of a high temperature, high pressure molten salt

methane pyrolysis reactor.

Total project cost: $2.3MLength 24 mo.

Binary Chloride Salts as Catalysts for Methane to

Hydrogen and Graphitic Powder

Total project cost: $1.2MLength 24 mo.

Production and continuous removal of graphitic powder from a molten salt methane pyrolysis

reactor.

Eric McFarland, C-ZeroAdditional team members in attendance: Zach Jones, Sam Shaner, Fadl Saadi

The Concept

DE-FOA-0001953 C-Zero LLC; McFarland Technical Volume 1953-1649

Page 1 of 16 AR-351-09.16

Contains Confidential, Proprietary, or Privileged Information Exempt from Public Disclosure

Molten-Salt Methane Pyrolysis Optimization Through in-situ Carbon Characterization and Reactor Design

C-Zero LLC (Santa Barbara, CA), Prof. Eric McFarland

High Value Methane Pyrolysis

$2.24 M (11% Cost Share) 24 Months

EXECUTIVE SUMMARY C-Zero is requesting assistance in its effort to commercialize a novel process for transforming methane into hydrogen and valorized carbon cement additive using high temperature liquids in a multi-phase pyrolysis reactor. ARPA-E support is requested to allow C-Zero to better understand and control the methane decomposition and carbon formation processes as well as model and demonstrate an optimized methane pyrolysis prototype. This will assist C-Zero in de-risking its process sufficiently to attract the additional private sector capital required for commercialization. In particular, the proposed work will allow C-Zero to: (1) monitor and control the molecular and morphological structure of the carbon formed during the methane pyrolysis process (2) identify reaction parameters that generate structural carbon that can be used as additives in cement (3) generate an experimentally-validated multi-physics model for a three-phase methane pyrolysis reactor.

NOTICE OF RESTRICTION ON DISCLOSURE AND USE OF DATA Pages [1 through 16] of this document may contain trade secrets or commercial or financial information that is privileged or confidential and exempt from public disclosure. Such information shall be used or disclosed only for evaluation purposes or in accordance with a financial assistance or loan agreement between the submitter and the Government. The Government may use or disclose any information that is not appropriately marked or otherwise restricted, regardless of source.

1

Decarbonize natural gas through Methane Pyrolysis

�

�

�

�

GoalIdea Key Challenges

Reactor Design

Monitor & Control Carbon Formation

Carbon Removal

Heat Integration

Carbon Valorization

Product Gas Cleanup

December 8, 2019

The Team

Zach JonesPresident & CEO

Prof. Eric McFarlandBoard Chair, CTO

Sam Shaner, Ph.D.Director of Engineering

Prof. Ches UphamUBC

Howard Fong, Ph.D.Chief Technical Strategist

Advisory Board

Arnie SmithFluor, Executive Director of Process Engineering

Philip GrossoKaiser Chemicals,

Former Vice President

Fadl Saadi, Ph.D.Director of Business

Development & Operations

Prof. Mike GordonUCSB

Prof. Roya MaboudianUCB

Collaborators

Prof. Horia MetiuUCSB

Prof. Raphaele ClementUCSB

2

Andrew Caldwell, Ph.D.Senior Scientist

December 8, 2019

Brett Parkinson, Ph.D.Senior Engineer

Lab & Setup

3December 8, 2019

Objectives for ARPA-E Project‣ Demonstrate in-situ spectroscopic measurements of carbon formation under

methane pyrolysis reaction conditions.

‣ Design and construct a 10 liter methane pyrolysis molten salt reactor with:– ≥ 70% CH4 conversion– ≥ 90% H2 selectivity– ≥ 5 mol H2/ m3 s– ≤ 2.5% wt salt in carbon product– High Pressure (≥ 5 bar)

4December 8, 2019

2020 2021

Begin Project G/NG: Demonstration of 10 liter reactor at

atmospheric pressure

Final Goal: Demonstration of 10 liter reactor at high pressure

Objectives for H2@Scale Project

5December 8, 2019

2020 2021

Begin Project G/NG: Demonstration of carbon removal system at atmospheric pressure

Final Goal: Demonstration of carbon removal system

at high pressure

‣ Demonstration of stable, active, binary chloride melt system:– ≥ 90% H2 selectivity– Graphitic carbon product that has properties favorable for battery anodes and

additives

‣ Design and construct a carbon removal system capable of:– High Temperature (1000 C)– Continuous carbon removal (≥ 24 hours)– High Pressure (≥ 10 bar)

Liquid Catalysts: Molten Metals

6December 8, 2019

2.E+00

0.00

0.02

0.04

0.06In Bi Sn Ga

Pure

Pb Ag

17%

Pt-

Sn17

% P

t-Bi

62%

Pt-

Bi

17%

Ni-I

n17

% N

i-Sn

73%

Ni-I

n17

% N

i in

Ga

17%

Ni i

n Pb

17%

Ni-B

i27

%N

i in

Au

27%

Ni-B

i27

% N

i in

Bi

17%

Cu

in S

n

Pb v

apor

Pure

Cu

Pt w

ire

Ni w

ire

TOF

- T

OF Bl

ank

(C

H 4 mol

ecul

es/a

tom

*s)

Liquid pure

Liquid

Liquid Ni alloys

Solid pure

metals

At solubility limit

Rel

ativ

e A

ctiv

ity

0.E+00

1.E-05

2.E-05

3.E-05

4.E-05

-1.25 -1.00 -0.75 -0.50 -0.25 0.00

Expe

rimen

tal a

ctiv

ity

(mol

H2

prod

uced

/ cm

2 *hr

)

Calculated Bader charge (on Pt or Ni)

Ni-Pb

Ni-Bi

Ni-Sn

Pt-Bi

Pt-Sn 1000 °C CH4 flowed over surface 17% Ni or Pt in all cases

Activity for: CH4 -> C + 2H2 Ni-

Bi!+e-Bi!+

Bi!+Bi!+Bi!+

Bi!+

e-e-e-

e- e-

‣ Liquid metal catalysts showed very high activity for methane pyrolysis‣ Ab initio molecular dynamics calculations support activation of host (BiàBi+) by the electrophile (NiàNi-)‣ Carbon purity and catalyst recovery costs were identified as technical and economic challenges

Science 358 (2017) 917-921

Liquid Catalysts: Simple Salts

7December 8, 2019

KCl NaBr NaCl KBrEa [kJ/mole] 280 270 300 330

‣ Monovalent salts have similar activities and produce similar disordered carbon‣ Experiments in excellent agreement with gas-phase, radical mediated kinetic pathway

Liquid Catalysts: Binary Salts

‣ Certain binary chloride salts show high catalytic activity for methane pyrolysis– Prolonged catalytic activity demonstrated (>24 hours)

‣ Activity of MnCl2-KCl correlates to tetrahedrally coordinated MnCl42- molecular ion‣ Surface mediated deep dehydrogenation demonstrated by isotope exchange

8December 8, 2019 Applied Catalysis B: Environmental 254 (2019) 659–666

concentration than the activation energies might suggest. In compar-ison to the pure KCl with an apparent activation energy of ˜300 kJ/mole, the reaction rate in MnCl2(50)-KCl(50) is only approximately 10times faster despite a factor of two reduction in apparent activationenergy. Interpretation of the decreased pre-exponential is difficultwithout a detailed model, however, one possible explanation is therelatively low surface area of the liquid resulting in a far lower collisionrate with the more active salt surface (or vapor phase salt) compared tothe total collision rate in the gas phase.

It was previously reported that a molten MnCl2-KCl mixture, con-tains two predominant molecular ions, MnCl42− and MnCl64-, and theirrelative ratios in the melt changes with composition [32–37]. ForMnCl2 molar fractions greater than 0.67, the concentration of MnCl42−

decreases and so does the activity of the melt [33]. Excess molar vo-lume and conductivity also showed the singular point at 67:33mol %mixture of KCl:MnCl2 [32]. One is tempted to suggest that MnCl42− ismore active catalytically than MnCl64-. However, CH4 is not highlysoluble in the melt and catalysis is controlled by the surface con-centration, which is not known.

MnCl2 has a high vapor pressure (0.13 atm at 1000 °C) which iscomparable to the partial pressure of the methane at the inlet (0.5 atm).It is difficult to determine whether the MnCl2 vapor is important or not.When the bubble first enters the reactor, it contains no MnCl2 vapor andit has the maximum methane partial pressure. When the bubble exitsthe melt, it has much less CH4 and some accumulated vapor producedby evaporation. Since we don’t know the evaporation rate, it is difficultto estimate the concentration of methane and of MnCl2 in the bubble, asthe bubble travels through the melt. In addition, there is vapor in thespace above the melt, even though this is purged with Ar. With the data

available to us, we cannot determine whether the MnCl2 vapor is anactive gas phase catalyst or not. Given the high vapor pressure of MnCl2and the relatively small pre-exponential for the most active mixtures itis likely reactions on the melt surface are the most important.

Fig. 2d shows the dependence of methane conversion on the flowrate at 1050 ℃. The flow rate determines the time in which the bubblegrows at the mouth of the tube through which the reactants are in-troduced. Once the bubble detaches from the tube its rise time throughthe melt depends on the depth, the bubble size, the melt density and itsviscosity. The bubble size and rise velocity are not a strong function ofthe flowrates [38]. The flow rates are such that the bubbles do notcoalesce as they rise through the column. To a first approximation,increasing the flow rate shortens the time the bubble grows at themouth of the inlet tube (and reacts). As expected, at the slower flowrates, the conversion increases as the time growing the bubble canbecome large compared to the rise time within the bubble. The decreaseis most significant when increasing from 5 sccm to 10 sccm.

In a continuous run of 30 h in KCl and in MnCl2(67)-KCl(33) me-thane conversion increases initially then stabilizes, Fig. 3. ForMnCl2(67)-KCl(33), at 1050 ℃ in the 12.5-cm molten salt bubblecolumn, the conversion increased from 45% to almost 55% and stabi-lized. Similar behavior was observed for pure KCl, except that thesteady state was reached after 5 h and the activity more than doubled,while still remaining well below that of the mixture MnCl2(67)-KCl(33).

In pure KCl the carbon produced in the KCl melt intermixes with thesalt forming a slurry. The increase of conversion in the first 5 h is likelydue to the accumulation of suspended carbon in the melt resulting in an

Fig. 2. Activity of molten MnCl2-KCl for CH4 pyrolysis. (a) CH4

conversion versus temperature for a MnCl2-KCl bubble column.Melt height 12.5 cm, reactant gas= 20 sccm (50mol% of CH4 and50mol% of Ar), sweep gas=50 sccm (100mol% of Ar). (b)Arrhenius plot for apparent activation energies for methane con-sumption rate, catalyzed by MnCl2-KCls. (c) Comparison of ap-parent activation energies and pre-exponential factors of themolten MnCl2-KCl mixtures. (d) CH4 conversion versus CH4 flowrate for a 12.5-cm molten MnCl2-KCl bubble column at 1050 ℃:reactant gas (100mol% of CH4), sweep gas= 50 sccm (100mol%of Ar).

Table 1Apparent activation energies of molten salts, metals and solid catalysts.

Catalyst Apparent activity energy (kJ/mole)

Carbon product

Molten MnCl2(50)-KCl(50) 153 ± 16 SeparableMolten MnCl2(67)-KCl(33) 161 ± 9 SeparableMolten Ni(27)-Bi(73) [23] 208 SeparableMolten Bi [23] 310 SeparableGas phase [39] 422 SeparableSolid Ni [40] 65 Non-separableSolid Ni/SiO2 [41] 96.1 Non-separableActivated carbon [40] 194 Non-separableCarbon black [42] 205-264 Non-separable

Fig. 3. CH4 conversion over 30 h of continuous running for the 12.5-cm moltenMnCl2(67)-KCl(33) and KCl(100) bubble columns at 1050 ℃: reactant gas=5sccm (100mol% of CH4), sweep gas=50 sccm (100mol% of Ar).

D. Kang, et al. $SSOLHG�&DWDO\VLV�%��(QYLURQPHQWDO���������������²���

���

Residence time ~ 1 s

Liquid Catalysts: Binary Salts

9December 8, 2019

to the presence of these complex ions.The pre-exponential factor of the rate constant in molten MnCl2-KCl

mixtures also decreases with increasing MnCl2 up to 50mol%, possiblysuggesting that the pyrolysis pathway within the gas bubble changes

from a predominately gas phase uncatalyzed reaction in the KCl melt toa reaction catalyzed by the melt surface or vapor due to the presence ofMnCl2. Considering overall compensation in the observed rates betweendecreasing activation energy and increasing pre-exponential factor, the

Fig. 5. TEM images of carbon extracted from MnCl2(67)-KCl(33) (a and b) and KCl(100) (e and f) with the insets showing the diffraction rings. HRTEM images ofcarbon extracted from MnCl2(67)-KCl(33) (c and d) and KCl(100) (g and h).

D. Kang, et al. $SSOLHG�&DWDO\VLV�%��(QYLURQPHQWDO���������������²���

���

MnCl2-KCl KCl

methane conversion in the molten MnCl2-KCl mixture was highest in amixture of 67:33.

The melt activity is stable for at least 30 h, producing a separablesolid carbon product due to the relatively high density of MnCl2-KClmelt and low wettability of the carbon product by a melt. The carbonproduced was graphitic and the salt contamination can be reduced toless than 7 at% with water washing.

The 99% H2 selectivity when using molten MnCl2-KCl suggests C2+

hydrocarbons are either not formed or are rapidly decomposed on thesurface. In contrast 10% selectivity to C2+ hydrocarbons was observedin pure KCl melt at the same CH4 conversion, 30%. The relatively largeamount of further deuterated methane, CH2D2, in H-D exchange onmolten MnCl2-KCl signifies that the reaction pathway is different thanthe assumed purely gas phase reaction in pure KCl melt, where CH3D isthe most abundant deuterated methane.

Considering the high CH4 conversion, H2 selectivity, and deuteratedmethane species seen using MnCl2-KCl melt, it can be surmised that theCH4 pyrolysis is accelerated and the graphitic carbon is generated onthe surface of the molten MnCl2-KCl. A single bubble model is in rea-sonable agreement with the experimental data within the molten saltbubble column at lab scales.

Acknowledgements

This work was primarily supported by the U.S. Department ofEnergy, Office of Science Basic Energy Sciences Grant number DE-FG03-89ER14048. We acknowledge the use of the Center for ScientificComputing supported by the California NanoSystems Institute and theMaterials Research Science and Engineering Center (MRSEC) at UCSanta Barbara through NSFDMR 1720256 and NSF CNS 1725797. TheMRL Shared Experimental Facilities are supported by the MRSECProgram of the NSF under Award No. DMR 1720256; a member of theNSF-funded Materials Research Facilities Network (www.mrfn.org).The authors are grateful for the indispensable technical assistance ofMr. Richard Bock of the UCSB Chemistry Department who prepared allthe quartz reactor components and made excellent suggestions as totheir modifications. All data are reported in the main paper and sup-plement.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in theonline version, at doi:https://doi.org/10.1016/j.apcatb.2019.05.026.

References

[1] A. Midilli, M. Ay, I. Dincer, M.A. Rosen, On hydrogen and hydrogen energy stra-tegies I: current status and needs, Renew. Sust. Energy Rev. 9 (2005) 255–271.

[2] A. Midilli, M. Ay, I. Dincer, M.A. Rosen, On hydrogen and hydrogen energy stra-tegies II: future projections affecting global stability and unrest renew, Sust. EnergyRev. 9 (2005) 273–287.

[3] N.Z. Muradov, T.N. Veziroglu, Green" path from fossil-based to hydrogen economy:an overview of carbon-neutral technologies, Int. J. Hydrogen Energy 33 (2008)6804–6839.

[4] J. Conti, P. Holtberg, J. Diefenderfer, A. LaRose, J.T. Turnure, L. Westfall,International Energy Outlook 2016 With Projections to 2040, ; USDOE EnergyInformation Administration (EIA), Office of Energy Analysis, Washington, DC(United States), 2016 pp. Medium: ED; Size: 290 p.

[5] F. Billaud, C. Gueret, J. Weill, Thermal-decomposition of pure methane at 1263-K -experiments and mechanistic modeling, Thermochim. Acta 211 (1992) 303–322.

[6] C.J. Chen, M.H. Back, R.A. Back, The thermal decomposition of methane. II.Secondary reactions, autocatalysis and carbon formation; non-arrhenius behaviourin the reaction of CH3 with ethane, Can. J. Chem. 54 (1976) 3175–3184.

[7] C.J. Chen, M.H. Back, R.A. Back, The thermal decomposition of methane. I. Kineticsof the primary decompositionto C2H6 + H2; Rate Constant for the homogeneousunimolecular dissociation of methane and its pressure dependence, Can. J. Chem.53 (1975) 3580–3590.

[8] A. Holmen, O. Olsvik, O.A. Rokstad, Pyrolysis of natural-gas - chemistry and processconcepts, Fuel Process. Technol. 249-267.

[9] M. Pudukudy, Z. Yaakob, M.Z. Mazuki, M.S. Takriff, S.S. Jahaya, One-pot sol-gelsynthesis of MgO nanoparticles supported nickel and iron catalysts for undilutedmethane decomposition into COx free hydrogen and nanocarbon, Appl. Catal. B:Environ. 218 (2017) 298–316.

[10] L. Zhou, L.R. Enakonda, M. Harb, Y. Saih, A. Aguilar-Tapia, S. Ould-Chikh, J.-l. Hazemann, J. Li, N. Wei, D. Gary, P. Del-Gallo, J.-M. Basset, Fe catalysts formethane decomposition to produce hydrogen and carbon nano materials, Appl.Catal. B: Environ. 208 (2017) 44–59.

[11] A. Rastegarpanah, M. Rezaei, F. Meshkani, H. Dai, H. Arandiyan, Thermocatalyticdecomposition of methane over mesoporous Ni/xMgO·Al2O3 nanocatalysts, Int. J.Hydrogen Energy 43 (2018) 15112–15123.

[12] A. Rastegarpanah, M. Rezaei, F. Meshkani, K. Zhang, X. Zhao, W. Pei, Y. Liu,J. Deng, H. Arandiyan, H. Dai, Influence of group VIB metals on activity of the Ni/MgO catalysts for methane decomposition, Appl. Catal. B: Environ. 248 (2019)515–525.

[13] K. Otsuka, S. Takenaka, H. Ohtsuki, Production of pure hydrogen by cyclic de-composition of methane and oxidative elimination of carbon nanofibers on

Fig. 6. (a) XRD patterns and (b) Raman spectra of carbon from MnCl2(67)-KCl(33) and KCl(100) reaction mixtures.

Fig. 7. (a) H2 and (b) C2+ hydrocarbon selectivities of the CH4 pyrolysis in the molten MnCl2(67)-KCl(33) and KCl(100) at 1050 ℃: reactant gas=10 sccm (100mol% of CH4), sweep gas=50 sccm (100mol% of Ar) (CH4 conversion was 30% in both cases).

D. Kang, et al. $SSOLHG�&DWDO\VLV�%��(QYLURQPHQWDO���������������²���

���

‣ High hydrogen selectivity (>98%) at 40% methane conversion

‣ Graphitization of carbon in MnCl2-KCl significantly higher compared to KCl– Indicative of carbon formation on the catalytic surface

‣ Moving Forward: Further investigation of binary chloride salts focused on identifying melt systems that are:– Highly catalytic– Produce graphitic carbon– Low cost and non-toxic

Applied Catalysis B: Environmental 254 (2019) 659–666

Direct Visualization of Carbon Deposition on Salt Surfaces

10December 8, 2019

100

mm

10 sccm CH4

Camera

TC

Carbon nucleation and growth on the surface of the

molten salt, 1000 C

Direct Visualization of Carbon Deposition on Salt Surfaces

11December 8, 2019

Camera

Trapped Single Bubble

‣ Insulating properties of molten salts allow for direct imaging of methane pyrolysis and carbon formation

‣ Moving Forward: Extend apparatus capabilities for in situ spectroscopic analyses of molten salt during methane pyrolysis by Raman, IR, and UV/Vis

– Help elucidate pyrolysis mechanisms for different systems

– Allow for rational design of reaction conditions for optimized hydrogen and specific carbon production

Reactor Modeling & Design

12December 8, 2019

‣ Constructed aqueous circulating fluid bubble column of sufficient size to avoid wall/entrance effects.‣ Evaluated multiple gas sparger

designs.‣ Moving Forward: Continue building

Eulerian multiphase CFD model to enable a verified predictive capability for building and scaling up methane pyrolysis bubbly flow reactor. Construct model system with continuous solid particulate removal capabilities.

Challenges and Potential Technical Partnerships‣ Biggest challenge to date has been the generation of a clean carbon product

– Metal contamination was a major issue early on– Moved to molten salt catalysts which have the advantage of:

• Lower catalyst cost• Ability to tune carbon morphology

‣ Another challenge is identifying suitable materials of construction (MOC)– Designing rapid MOC testing of several different potential materials.

‣ Looking for technical partnerships with entities that have expertise in carbon modification and characterization

13December 8, 2019

T2M‣ C-Zero is working on the design and commercialization

of its methane pyrolysis technology ‣ C-Zero plans to complete the full design of the

commercial reactor by 2022‣We are looking for partnerships/collaborations with:

– Current consumers of natural gas– Experts in delayed cokers– Hydrogen producers and consumers– Refineries (especially in California)

14December 8, 2019

�

�

�

�